Criteria

What it is

Criteria (legal name Criteria Corp) sells pre-hire assessments and structured interviewing to mid-market employers. The assessment side is the older half: the CCAT cognitive aptitude test, personality and risk inventories, skills tests, an emotional-intelligence test, and game-based cognitive tests. The interviewing side is newer: one-way async video, a real-time video interview tool, AI scoring of interview transcripts, and, since June 2026, an AI Interview Agent that runs a structured first-round interview by voice or chat.

Two acquisitions explain the catalog. Criteria bought Revelian, the Brisbane assessment company behind its game-based and emotional-intelligence tests, in 2020, and Alcami Interactive, an on-demand video interviewing platform, in 2021. Revelian is why Criteria runs a second headquarters in Brisbane next to its West Hollywood office. Founder Josh Millet has been CEO since 2006; Sumeru Equity Partners took a majority growth stake in 2019 after Level Equity’s 2015 Series A.

Scale figures are the vendor’s own: 4,000+ customer organizations in 60 countries and 80 million+ assessments delivered. In 2026 it reported ISO 42001 certification (the AI management-system standard) across its AI products.

Criteria is not an ATS, a sourcing tool, or an autonomous AI recruiter. It plugs into the ATS you already run and scores candidates at one or two stages of the funnel.

Why it shows up in Recruiting stacks

• The CCAT is the default cognitive screen for mid-market hiring. It is a 50-question, 15-minute test, short enough for volume roles and backed by two decades of validation data. When a hiring manager asks for “the Criteria test,” this is usually the one they mean.
• Flat subscription, uncapped testing. Criteria sells annual plans with unlimited assessments rather than per-candidate credits. For a contact center or sales floor testing thousands of applicants a year, the unit cost falls with every candidate you add.
• Assessments and interviews in one contract. The 2026 plan structure makes this explicit: Criteria Assess (all test types, TestMaker, standard proctoring), Criteria Interview (async and real-time video with transcripts and summaries), and Criteria Hire (both). Each has an “AI” upgrade that adds AI proctoring, AI interview scoring, and the AI Interview Agent.
• The integration list is long. 75+ technology partners, including Greenhouse, iCIMS, SmartRecruiters, Jobvite, Workday, Lever, Ashby, Workable, JazzHR, Bullhorn, UKG Pro, and Oracle Taleo, plus a HackerRank partnership for coding tests.

Pricing

• Quote only. The plans page lists Assess, Interview, and Hire but no prices; every path ends at “Request Pricing.” A free trial is offered.
• Real-world band (third-party estimates, not vendor-confirmed): 2026 buyer guides put small-business contracts at roughly $5,000 to $15,000 a year, mid-market at $15,000 to $30,000, and enterprise above $40,000. Effective cost per assessment lands around $30 to $60 on starter plans and $10 to $20 at enterprise volume. Several of those guides come from competitors (Compono publishes one), so treat the band as a negotiating anchor, not a quote.
• Contract shape: 12-month minimum, no month-to-month. You pay the full year through a hiring freeze.
• Add-ons that move the price: AI Proctoring, AI features, the AI Interview Agent, premium integrations, and Criteria Develop (the post-hire module with manager check-ins, TEAMscan, and an AI coach).
• Anchor for comparison: TestGorilla’s 2026 Core plan runs about $2,580 a year for 750 candidate credits, roughly $3.40 per tested candidate. If your volume is under a few hundred candidates a year, Criteria has to justify a multiple of that on validity and interviewing, not on test count.

Best for

Mid-market TA teams (roughly 200 to 5,000 employees) that hire at volume into contact center, sales, customer service, and early-career roles, want a validated cognitive and personality screen at the top of the funnel, and would rather buy assessments and structured video interviews from one vendor than two. It fits best where the ATS is on the partner list and the team tests at least several hundred candidates a year, enough to make an unlimited-use subscription cheaper than credits.

Do not buy Criteria if you hire fewer than about 100 people a year, you mainly need coding assessments, or you need the deepest defensibility for regulated or senior-leadership selection.

Versus the alternatives

• SHL: the market-share leader in enterprise psychometrics. Pick it when the selection decision has to survive a legal challenge across many countries, and budget accordingly. Criteria wins on price, setup speed, and the plain-English reports mid-market hiring managers actually read.
• HireVue: the share leader in enterprise video interviewing, with assessments attached. Pick it for Fortune 500 volume programs that need interviewing at global scale. Criteria wins when the cognitive test is the core of the screen and the interview is a second step.
• TestGorilla: the fastest-growing entrant, with self-serve plans and a 400+ test library. Pick it for SMB budgets, many role-specific skills tests, and a free tier to start. Criteria wins on validation depth for cognitive and personality tests and on the unlimited-volume model at scale.
• Harver: high-volume job simulations and matching. Pick it for frontline and hourly programs that hire thousands a month and want realistic job previews.
• Vervoe: AI-graded work-sample tests. Pick it when you want candidates doing the job rather than answering aptitude items.

If none fit because volume is low, a written work sample scored against a rubric, followed by a structured interview, beats any subscription below about 100 hires a year. For the broader shift, see skills-based hiring.

Watch-outs

• The AI Interview Agent is new. The June 16, 2026 release announced it, while the interview product page still labels it “coming soon.” Guard: ask for a live demo on your own questions and a reference customer already running it in production, and keep it out of the contract’s committed scope until you have seen it work.
• CCAT prep is everywhere, and so is AI-assisted test-taking. Practice tests for the CCAT are a cottage industry, and candidates use second devices and chatbots. Standard proctoring does not catch that; AI Proctoring samples photos, screen, and audio at intervals rather than streaming continuously. Guard: turn on AI Proctoring for remote, unsupervised testing, and re-verify high scorers with a short supervised item set or an interview question that probes the same reasoning. See AI interview fraud.
• You own the automated decision. Cutoff scores that auto-reject candidates, and AI scoring of interviews, fall under NYC Local Law 144 bias-audit rules, the Illinois AI Video Interview Act for video, and the EU AI Act’s high-risk category. ISO 42001 certifies Criteria’s management system, not your cutoffs. Guard: run an adverse-impact check by segment on your own pass rates every quarter (the screening score variance audit is built for this) and keep a human on every reject.
• Annual lock-in with no usage floor. The 12-month term with no monthly option hurts seasonal and low-volume hirers. Guard: negotiate a mid-term downgrade right, or a price cap on renewal, before signing. Capterra reviewers already flag rising renewal prices.
• Integration depth varies by ATS. Reviewers report Workday and Indeed Easy Apply flows needing workarounds. Guard: before signing, test the full loop in a sandbox (trigger, invite, score write-back to the candidate profile) on your ATS, not on the vendor’s demo tenant.
• Long batteries cost completions. Stacking CCAT, personality, and a skills test can push a first-stage screen well past half an hour, and reviewers tie length to drop-off. Guard: cap the first-stage battery at about 20 minutes and move the rest after a first conversation.